Grades are a hygiene check, not a security guarantee.

MCP, MCP server

Agent Guardrail MCP Server

Agent Guardrail MCP Server, an MCP server listed in the PlugVet directory.

Compatibility: Not stated.

Grade B

Minor issues only, no Medium or High. Worth a quick look at the findings below before installing.

Graded by PlugVetRubric v2

Not a guarantee. A hygiene check of the source at the commit shown, not a security audit.

Rescanned at commit 9f74ba7d4f31, unchanged since the check below.

NETWORK_NO_FIXED_HOST SECRETS_PATH_UNCONFIRMED PERMISSION_SHELL_DECLARED SECRETS_ENV_READ

Why this grade

  • NETWORK_NO_FIXED_HOST Low

    Makes a network call whose host is not written as a fixed address.

    mcp_guardrail_http.py, line 62 with urllib.request.urlopen(req) as resp:

  • SECRETS_PATH_UNCONFIRMED Low

    The Secrets check reports a credential or environment file, but no quoted line shows a recognisable path.

    mcp_guardrail_http.py, line 29 GUARDRAIL_API_KEY = os.environ.get("GUARDRAIL_API_KEY", "")

  • PERMISSION_SHELL_DECLARED Note

    Runs shell commands, and the manifest or description says so.

    examples/decorator.py, line 13 return subprocess.run(cmd, shell=True, capture_output=True).stdout

  • SECRETS_ENV_READ Note

    Reads an environment variable whose name contains KEY, SECRET, TOKEN, PASSWORD or CREDENTIAL.

    mcp_guardrail_http.py, line 29 GUARDRAIL_API_KEY = os.environ.get("GUARDRAIL_API_KEY", "")

Any High, or two or more Medium findings, give D. One Medium gives C. Low findings or a security Note give B. Each code names a finding behind the letter.

Grades are a hygiene check of the source at the commit shown, on the date checked, not a security audit or a guarantee. The free directory grades are automated. Before release, a PlugVet engineer reviews every paid report and any grade we hold back, including all F grades, against the source code and the evidence we quote.

What was checked

Plain notes from the hygiene checks.

  • Permissions

    Note

    requests shell access. The manifest names this permission.

    PERMISSION_SHELL_DECLARED

    examples/decorator.py, line 13

    return subprocess.run(cmd, shell=True, capture_output=True).stdout

  • Install and update

    Pass

    No install-time script found, and dependencies are pinned or absent.

  • Network

    Concern

    can reach a host that is not written as a fixed address.

    NETWORK_NO_FIXED_HOST

    mcp_guardrail_http.py, line 62

    with urllib.request.urlopen(req) as resp:

  • Secrets

    Concern

    reads a credential or environment file.

    SECRETS_ENV_READ SECRETS_PATH_UNCONFIRMED

    mcp_guardrail_http.py, line 29

    GUARDRAIL_API_KEY = os.environ.get("GUARDRAIL_API_KEY", "")

    agent_guardrail/billing.py, line 82

    self._api_key = os.environ.get("BLOCKONOMICS_API_KEY", "").strip()

    agent_guardrail/billing.py, line 83

    self._webhook_secret = os.environ.get("BLOCKONOMICS_WEBHOOK_SECRET", "").strip()

    Show more

    agent_guardrail/billing.py, line 398

    return bool(os.environ.get("STRIPE_SECRET_KEY", "").strip())

    agent_guardrail/billing.py, line 424

    secret_key = os.environ.get("STRIPE_SECRET_KEY", "").strip()

    agent_guardrail/billing.py, line 486

    secret_key = os.environ.get("STRIPE_SECRET_KEY", "").strip()

    agent_guardrail/billing.py, line 487

    webhook_secret = os.environ.get("STRIPE_WEBHOOK_SECRET", "").strip()

    agent_guardrail/policy.py, line 12

    "target_denylist": ["/etc/*", "/root/*", "*.env"],

    agent_guardrail/policy.py, line 303

    "target_denylist": ["/etc/*", "/root/*", "*.env", "*.key", "*.pem"],

    agent_guardrail/policy.py, line 317

    "*.env",

    agent_guardrail/policy.py, line 334

    "target_denylist": ["*.env", "*.key", "*.pem"],

    agent_guardrail/proxy.py, line 143

    _billing_api_key = os.environ.get("BLOCKONOMICS_API_KEY", "").strip()

    .gitignore, line 16

    env/

    README.md, line 240

    "target_denylist": ["/etc/*", "*.env", "*.key"], # These targets always denied

    docs/nlnet-application.md, line 42

    5. Target denylist matching - deny if the target path or URL matches a forbidden pattern (e.g., `/etc/*`, `*.env`, `*credentials*`)

  • Agent instructions

    Pass

    No instruction to ignore earlier instructions or to hide a step.

  • Obfuscation

    Pass

    No encoded or minified payload found.

  • Licence

    Pass

    A licence file is present (LICENSE).

  • Maintainer activity

    Pass

    The latest commit is within the last 365 days.

Note

Hygiene check of commit 9f74ba7d4f31. requests shell access. The manifest names this permission. can reach a host that is not written as a fixed address. reads a credential or environment file.

Date and source

Date checked:

View source

github.com

Commit graded: 9f74ba7d4f31daa06aae9fe374d6414e88372573

This link goes to the original host. PlugVet does not host or check what you download from there, and installing it is your own decision and risk.

Read how this letter is chosen